OCR Image AI
Last updated
Last updated
The OCR Image AI feature allows AutoForward to automatically extract text from any image and include it as part of the forwarded message. This is ideal for screenshots, flyers, or images containing important written information.
Select the Task you want to apply AI MODE to
Select the AI MODE option inside the bot
Select OCR Image AI
4. Initial Screen (Feature Disabled)
✅ Model: You must first select an AI model (e.g., gpt-4o, gemini-pro)
Toggle: The “OCR Image AI” toggle is off by default.
⚠️ You must choose a model that supports image input and text extraction.
Toggle the switch to enable/disable OCR functionality.
When enabled, AutoForward will process incoming images through the selected model and extract text.
💡 Note: Using AI may introduce slight delays in forwarding due to image processing time.
You can choose how the AI interprets and formats the extracted content:
System Prompt (default)
A pre-configured prompt optimized for clean and actionable OCR output.
Custom Prompt
Define your own instruction, such as:
“Extract only numbers and email addresses from the image.”
“Summarize the main points in bullet form.”
Use this to tailor the AI’s output to your use case.
Select where the extracted text will appear in the forwarded message:
Prepend to Caption
Adds the OCR result before the existing message content.
Example:
(Future options may include: Append to Caption, Replace Message Content, etc.)
Must have a valid OpenAI API key with access to vision models.
Your chosen model must support both image inputs and textual outputs.
